Arsenic Transfer and Biotransformation in a Fully Characterized Freshwater Food Web

Coordination Chemistry Reviews 306 (2016) 558–565

Contents lists available at ScienceDirect

Coordination Chemistry Reviews

j ournal homepage: www.elsevier.com/locate/ccr

Review

Arsenic transfer and biotransformation in a fully characterized freshwater food web

a,∗ a a,1 b

Richard D. Foust Jr. , Anne-Marie Bauer , Molly Costanza-Robinson , Dean W. Blinn ,

c d,2 d,2

Roger C. Prince , Ingrid J. Pickering , Graham N. George

Department of Chemistry and Biochemistry, Northern Arizona University, Flagstaff, AZ 86011-5698, USA

Department of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011-5460, USA

ExxonMobil Research and Engineering Company, Route 22 East, Annadale, NJ 08801 USA

Stanford Synchrotron Radiation Laboratory, Menlo Park, CA, USA

Contents

1. Introduction ...... 559

1.1. Environmental transformations of arsenic ...... 559

1.2. Chemistry of arsenic in food webs ...... 560

1.3. Montezuma Well ecosystem ...... 560

2. Sample collection and total arsenic analysis...... 561

2.1. XANES data collection and analysis ...... 561

3. Results and discussion ...... 562

4. Conclusions ...... 564

Acknowledgments ...... 564

References ...... 564

a r t i c l e i n f o a b s t r a c t

Article history: X-ray absorption near edge spectroscopy (XANES) was combined with ICP-MS to understand arsenic

Received 1 December 2014

transfer and transformation within the freshwater Montezuma Well (central Arizona, USA) food web.

Received in revised form 22 February 2015 −1

Montezuma Well water contains 110 ␮g L arsenic (100% H3AsO4), which was shown previously to

Accepted 6 March 2015

originate from ore deposits approximately 30 km to the west, and transported underground to enter

Available online 16 March 2015

Montezuma Well through vents in the bottom of this collapsed travertine spring. The Montezuma Well

food web contains three trophic levels with only ﬁve organisms in the top two levels, making it possible

Keywords:

to account for the arsenic in >90% of the biomass of the food web. Arsenic diminution generally occurs

Arsenic −1 −1

Biotransformation between trophic levels, with 702 mg kg As in the primary producers, 3.4 mg kg As in the second

−1

Diminution trophic level, and <1.3 mg kg As for the top-tier of the littoral zone food web. A notable exception to

−1

Food web the biodiminution trend is the very high total arsenic content (2810 mg kg ) of Motobdella montezuma,

Montezuma Well an endemic leech and top predator. The biotransformed (sulfur-coordinated) arsenic in M. montezuma

appears to be present almost entirely on the surface of the organism, possibly suggesting a detoxiﬁcation

mechanism. XANES sample spectra were ﬁt by a linear combination of model arsenic compound spectra

and indicated that arsenic enters the food web from the Well water entirely as inorganic As(V) and

is transformed within the food web into more reduced and organic arsenic species, including sulfur-

coordinated and methylated compounds.

∗

Corresponding author. Current address: Department of Chemistry and Biochemistry, James Madison University, Harrisonburg, VA 22807, USA. Tel.: +1 011 540 568 6800;

fax: +1 011 540 568 7938.

E-mail address: [email protected] (R.D. Foust Jr.).

Current address: Program of Environmental Studies and Department of Chemistry and Biochemistry, Middlebury College, Middlebury, VT 05753, USA.

Current address: Department of Geological Sciences, University of Saskatchewan, Saskatoon, SK S7N 5E2, Canada.

http://dx.doi.org/10.1016/j.ccr.2015.03.005

R.D. Foust Jr. et al. / Coordination Chemistry Reviews 306 (2016) 558–565 559

Table 1

1. Introduction

Chemical forms of arsenic substances found in aquatic environments, their formulas

and abbreviations [22,23].

Arsenic deserves a special place among the chemical elements

Name Formula Abbreviation

because of its chemistry and its interesting history. Arsenic has been

known as a substance since antiquity and was named “arsenikon” 1 Arsenous acid or arsenite H3AsO3 As(III)

by the Greeks, their name for the yellow mineral orpiment, As2S3. 2 Arsenic acid or arsenate H3AsO4 As(V)

3 Monomethylarsonous acid CH3As(OH)2 MMA(III)

Arsenic was identiﬁed as a chemical element by Albertus Magnus

4 Monomethylarsonic acid CH3AsO(OH)2 MMA(V)

in the 1200s [1]. Arsenic is the 20th most abundant element of the

5 Dimethylarsinous acid (CH3)2As(OH) DMA(III)

Earth’s crust and over 245 arsenic bearing minerals are known [2].

6 Dimethylarsinic acid (CH3)2AsO(OH) DMA(V)

Arsenic trioxide is an important industrial chemical. The United 7 Trimethylarsine (CH3)3As TMA

8 Trimethylarsine oxide (CH ) AsO TMAO

States imported more than 20,000 tons of arsenic trioxide annually 3 3

9 Arsenocholine (CH3)3As CH2CH2OH AsC

during 2001–03. This quantity dropped to 6100 tons when the use + −

10 Arsenobetaine (CH3)3As CH2CO2 AsB

of chromated copper arsenate as a wood preservative was banned. a

11 Arsenosugars (CH3)2AsOCH2C4H4O–R AsS

High-purity arsenic (99.9999%) is used in the electronics industry 12 Arsenic glutathione complex (GS)3As(III) (GS)3As(III)

to manufacture semi-conductors, solar cells, and specially coated a

R can be an alcohol, phosphate, sulfate, sulﬁte, carboxylic acid or a heterocycle.

optical materials [3]. b

GS represents the glutathione anion.

Hippocrates (460–357 BC) used orpiment (As2S3) and realgar

(As2S2) as a corrosive salve and to treat ulcers [4]. Pliny the Elder

were discovered in apples as early as 1919 when experiments

(23–79 AD) described the preparation of an arsenic liniment made

were begun to identify washing procedures to eliminate the toxic

with vinegar that was suggested as a remedy for asthma, cough,

residues. Although lead remains in the top 5–20 cm of soil, the

and several other ailments [1]. Modern medicinal uses of arsenic

arsenic may be transported through runoff or through groundwa-

include using sodium cacodylate and sodium arsanilate to treat pel-

ter because of its higher solubility. In soluble forms, arsenic can be

lagra, malaria, and sleeping sickness [1,4,5]. In 1909, arsphenamine

transported through the root system into plants or to wells where

was proposed by Paul Ehrlich as a treatment for syphilis and try-

they can be ingested through drinking water. Transport via ground-

panosomiais. This practice continued for nearly 40 years until

water has resulted in millions of people living in the Bengal Delta

penicillin became available [4]. Arsphenamine is considered the

Plain in Bangladesh and the West Bengal area of India to acquire

ﬁrst chemotherapeutic agent [6].

arsenic poisoning from their drinking water, although in this case

The toxic properties of arsenic were reported by Aristotle in

the underlying source of arsenic in the water is geologic, rather than

340 BC when he wrote that realgar (As2S2) kills horses and cat-

human activity [11–13].

tle when put into water [1,7]. Arsenic’s poisonous properties have

Rice accumulates arsenic at higher levels than other terrestrial

been used to change the course of history. Nero is reported to have

plants. Indeed, dietary arsenic from rice grown in contaminated

poisoned Britannicus in 55 AD to secure his Roman throne, and Pope

ﬁelds is known to be a contributing factor for arsenic poisoning in

Alexander IV, and his son, Cesare Borgia, are legendary for their use

the Bengal Delta [14,15]. The American Chemical Society addressed

of arsenic solutions to kill enemies [4]. The Chinese are credited

this concern in a recent symposium on arsenic contamination in

with using arsenic sulﬁdes as an insecticide as early as 900 AD, and

food and water [16], and the U.S. Department of Agriculture has a

arsenic oxide was reportedly used in ant bait in Europe in 1699 [8].

program to measure the arsenic content of rice grown in the United

The widespread global use of arsenical insecticides over the past

States [17].

150 years represents one of society’s major environmental catas-

It is essential to understand the environmental chemistry of

trophes. Copper acetoarsenite, commonly known as Paris Green,

arsenic in order to develop methods for remediating contaminated

was ﬁrst used in 1867 on the Colorado potato beetle in the USA.

arsenic soil and water. The solubility of arsenic, and therefore its

Paris Green sprays were soon used by fruit growers to control the

ability to be transported in an aqueous environment, is governed

codling moth (Cydia pomonella), an insect pest capable of destroy-

by the oxidation state of arsenic, the presence of other metals in

ing up to 98% of a farmer’s apple production [8,9]. Paris Green

the soil or aquifer material and their oxidation states [2,18–20].

was also used internationally in mosquito abatement programs

The toxicity of arsenic also depends on the oxidation state and on

for which it was applied directly to water bodies as a powder.

the ligands that surround the arsenic atom, making the chemical

Lead arsenate was ﬁrst used as an insecticide against the gypsy

form of the arsenic as important as the amount of arsenic when

moth (Lymantria dispar) in 1892 in Massachusetts, USA. Lead arse-

assessing environmental hazards [21]. Until recently, the litera-

nate proved to be more effective than Paris Green sprays for the

ture on environmental arsenic has emphasized arsenic quantity

gypsy moth because of its lower solubility in water and its ten-

present in environmental samples and has paid less attention to

dency to adhere to the surfaces of plants. Lead arsenate became the

its chemical form.

insecticide of choice for many agricultural crops including apples,

cherries, cotton and many vegetables [8–10]. Lead arsenate use

increased steadily in the United States, peaking in 1944 at 86.4 1.1. Environmental transformations of arsenic

million pounds annually. Sophisticated sprayers were designed to

apply lead arsenate over large areas of farmland and orchards. More than 50 organoarsenic compounds are found in nature

Adding lead arsenate to irrigation water became a common practice [22,23]. Table 1 lists the most common arsenic substances found

in some locations [9]. in aquatic environments. Inorganic arsenic species predominate in

The discovery of DDT’s insecticidal properties in 1947 provided aqueous environments, and depending upon the redox properties

a convenient alternative to lead arsenate. Lead arsenate insecticides of the system, the arsenic may be present in the +5 oxidation state

were quickly replaced by DDT, largely because the insect pests had as arsenic acid (H3AsO4) or in the +3 oxidation state as arsenous

built up resistance to lead arsenate and DDT was more effective acid (H3AsO3) [20,21]. Most arsenic compounds isolated from bio-

at lower doses. All insecticidal uses of lead arsenate in the United logical organisms are organic species that are produced through

States were ofﬁcially banned in 1988 [8]. metabolic processes in phytoplankton, bacteria or in the organs

Unfortunately, hundreds of thousands of hectares of contam- of vertebrates [21,24,25]. Two recent review articles document

inated soil are the legacy that remains from a century of relying arsenic concentrations in an extensive list of organisms and the

on arsenic-based insecticides. Elevated levels of arsenic and lead species of arsenic that have been identiﬁed in these organisms

560 R.D. Foust Jr. et al. / Coordination Chemistry Reviews 306 (2016) 558–565

Table 2

[21,26]. The transformation of inorganic arsenic to organoarsenic

−1

Chemical characteristics of Montezuma Well water (concentrations in mg L ,

substances proceeds via a series of biomethylation steps involv- −1

except water As concentration, which is expressed in ␮g L and sediment As con-

ing S-adenosylmethionine, following a pathway ﬁrst proposed by −1

centration, which is expressed as mg kg dry wt) [39,41].

Frederick Challenger [21,27–29].

Ca 117.0 ± 4.5

Mammals metabolize inorganic arsenic via methylation in the 2+

Mg 37.4 ± 2.2

liver to methylarsonic acid, MMA(V), and dimethylarsinic acid, +

Na 50.6 ± 0.2

DMA(V) [30]. DMA(V) and MMA(V), the ﬁnal arsenic metabolic K 5.10 ± 0.02

2−

products in mammals, are excreted from the body through urine. SO4 10.8 0.8

−

Cl 44.6 ± 3.4

Methylation of arsenic was long considered a detoxiﬁcation pro- −

HCO3 595.0 ± 5.0

cess for this reason [22]. However, reduced forms of the methylated

SiO2 21.9 ± 3.0

metabolites, MMA(III) and DMA(III), are highly toxic and have been

As (water) 110

observed in urine [22,30]. As (sediment) 427

CO 500

Freshwater phytoplankton convert inorganic arsenic into the 2

O2 (daytime) >12

less toxic DMA(V), which is then excreted from the organism [31].

O2 (nighttime) <4

Biotransformation of inorganic arsenic to organoarsenicals is there-

pH 6.5 ± 0.02

fore, considered the main detoxiﬁcation/defense mechanism of

phytoplankton [26]. substances, but they are not typically found

−1

in the aqueous environment. Arsine and methylated arsines are in Grace Lake to 340 mg kg in Kam Lake. Only inorganic and

produced by fungi in reducing environments [32]. predominantly oxidized forms of arsenic were observed within

the phytoplankton. Zooplankton contained smaller percentages of

1.2. Chemistry of arsenic in food webs inorganic arsenic, ranging from 38% in Grace Lake to 98% in Kam

Lake. DMA(V) and MMA(V) were observed in zooplankton at levels

Giddings and Eddlemon were the ﬁrst to demonstrate the move- ranging from 0.1 to 2%. Arsenosugars were observed in zooplankton

ment of arsenic between trophic levels [33]. Using arsenic-74, samples from all three lakes, but Grace Lake, with the lowest ambi-

a radioactive isotope, to follow arsenic concentrations, bioaccu- ent arsenic concentration, contained a greater variety and higher

mulation factors of 965, 192 and 11 were reported for the ﬁrst relative abundance of arsenosugars. Arsenobetaine was observed

(phytoplankton), second (zooplankton), and third (snail) trophic in zooplankton from Grace Lake, but was absent in samples from

levels, respectively. Lindsay and Sanders [34] similarly observed Kam Lake and Long Lake.

signiﬁcant arsenic uptake by phytoplankton in an estuarine sys-

tem, but only a slight increase in arsenic concentrations in the 1.3. Montezuma Well ecosystem

grass shrimp that feed on the phytoplankton. Further examining the

seasonality of arsenic uptake, Chen and Folt [35] found that arsenic Montezuma Well, a collapsed travertine spring, is located in cen-

◦

concentrations in the freshwater lake phytoplankton peaked dur- tral Arizona, USA, in an upper Sonoran Desert grassland (34 39 N,

−1 ◦

ing the summer at 25 mg kg . Arsenic concentrations diminished 111 45 W). Water enters the Well through four large vents in the

3 −1

with each successive level in the four-step food chain, with arsenic ﬂoor of the spring (4150–7300 m day ) and exits at the south-

−1

concentrations in ﬁsh of <1 mg kg . east corner, where it passes through a limestone cave system to

Maeda et al. improved our understanding of arsenic metabolism eventually merge with Wet Beaver Creek [39]. The feed water is

in freshwater food chains by adding arsenic speciation informa- supersaturated with carbon dioxide and outgassing occurs at the

tion. In a study of a three-trophic level system (an autotroph bottom of the Well, resulting in dissolved CO2 levels exceeding

−1

(alga, Chlorella sp.), a grazer (zooplankton, Moina sp.) and a car- 500 mg L . The feed water contains dissolved arsenic transported

nivore (goldﬁsh, Carassius sp.) [36], water arsenic levels of 30 underground over a distance of 30 km, resulting in an arsenic con-

−1 −1

and 100 mg L resulted in alga arsenic concentrations of 745 and centration of 110 ␮g L [40]. The source of arsenic in Montezuma

−1

2850 mg kg , respectively. More than 98% of the algae arsenic well is ore deposits located near Jerome, AZ on the west side of the

was found to be non-methylated inorganic. While the relative con- Verde Valley [41]. The water surface of the Well is circular with a

centration of methylated arsenic increased successively moving up diameter of 112 m, and the maximum depth, at the center of the

the three-step freshwater food chain, the total arsenic accumula- pool, is 17 m. Water chemistry of Montezuma Well is summarized

tion decreased by an order of magnitude with each trophic level. in Table 2.

The authors also demonstrated that the bioaccumulated arsenic Exceedingly high levels of CO2, abundant sunshine, and unusu-

came from both water and food in the two higher trophic levels ally constant water temperature year round within Montezuma

by raising the zooplankton and carnivores in water that contained Well result in some of the highest primary productivity numbers

different concentrations of arsenic. The trends of biodiminution reported [42]. On an annual basis, 1698 g-carbon (dry weight) is

and increased relative abundance of methylated arsenic species produced by photosynthetic activity for each m of Montezuma

with trophic level was observed in another three-step food chain Well [43]. The unusual water chemistry and the isolated location

consisting of green alga (Chlorella vulgaris), shrimp (Neocaridina have resulted in an ecosystem with a high proportion of endemic

denticulate) and killiﬁsh (Oryzias latipes) [37]. In fact, methylated species. The dominant organisms in the Montezuma Well food web

arsenic compounds became the dominant form of arsenic in the are listed in Table 3.

second and third trophic levels. A striking feature of the food web is the small number of orga-

Arsenic transfer between phytoplankton and zooplankton and nisms that are involved. Sixty-ﬁve percent of the phytoplankton

changes in speciation in three Canadian lakes was studied by community consists of only two species (Coccomyxa minor and

Caumette et al. [38]. This study is noteworthy because of the highly Nannochloris bacillaris) and only ﬁve species (Ankistrodesmus sp.,

−1

variable arsenic concentrations in each lake (7–250 ␮g L ), due to Chlorella sp., Monoraphidium sp., Nephrocytium sp. and Scenedesmus

historic mining activities in the area. Phytoplankton arsenic lev- sp.) make up the next 12% of the phytoplankton assemblage [41].

−1

els ranged from a low of 154 mg kg in Grace Lake to a high of The zooplankton community is dominated by Hyalella montezuma,

−1

894 mg kg in Kam Lake, proportional to arsenic concentrations in an endemic freshwater amphipod. H. montezuma is so critical to

the water. Similarly, zooplankton arsenic levels varied in propor- the viability of the Well’s food web that it is labeled a keystone

−1

tion to water and phytoplankton concentrations, from 7 mg kg organism. The highest trophic level consists of only four insects

R.D. Foust Jr. et al. / Coordination Chemistry Reviews 306 (2016) 558–565 561

Table 3

Dominant organisms of the Montezuma Well ecosystem [41].

Primary producers

Phytoplankton Coccomyxa minor Cyanobacteria Chroococcus

Phytoplankton Nannochloris bacillaris Cyanobacteria Gloeocapsa

Phytoplankton Ankistrodesmus sp. Cyanobacteria Merismopedia

Phytoplankton Chlorella sp. Cyanobacteria Oscillatoria

Phytoplankton Monoraphidium sp. Diatoms Achnanthidium

Phytoplankton Nephrocytium sp. Diatoms Cocconeis

Phytoplankton Scenedesmus sp. Diatoms Gomphonema

Zooplankton

Amphipod Hyalella montezuma Amphipod Hyalella azteca

Copepod Tropocyclops prasinus mexicanus

Insects

Water scorpion Ranatra montezuma Damsel ﬂy Telebasis salva

Giant water bug Belostoma bakeri Dytiscid beetle Cybister ellipticus

Other invertebrates

Leech Motobdella montezuma

in the littoral zone (Ranatra montezuma, Belostoma bakeri, Telebasis net and manually removing the material collected with a stainless

salva and Cybister ellipticus) and an endemic leech (Motobdella mon- steel spatula and placing it into centrifuge tubes. Sediment samples

tezuma) in the limnetic zone. The four insects listed constitute over were collected with a plastic scoop, placed in Ziplock bags, sealed

80% of the insect biomass at Montezuma Well. Half of the organisms and placed on ice for transport to the laboratory. All samples were

◦

in the Well’s food web are endemic [43]. stored at −80 C until analyzed.

Freshwater amphipods commonly feed on dead organic mat- Total arsenic in samples was determined by ICP-MS after diges-

ter and associated microbes. H. montezuma is unique among North tion in a Milestone Ethos closed-vessel, temperature-controlled

American amphipods, being the sole planktonic ﬁlter feeder. H. microwave digestion system. Approximately 0.5 g of sample was

montezuma swims freely in the water column, feeding on phyto- weighed into Teﬂon digestion vessels, into which 9.0 mL of HNO3

plankton by ﬁltering an average of 15.6 mL of water a day. The four and 3.0 mL of H2O2 were added. Each digestion set of 12 samples

insects and the endemic leech, M. montezuma, feed exclusively on H. contained 2–4 SRMs, 6–8 samples and two digestion blanks

◦

montezuma [43]. These feeding behaviors make it possible to study (reagents only). Sample vessels were heated up to 180 C over

◦

the movement of arsenic through the entire food web. 5.5 min and held at 180 C for 11 min at 600 W, allowed to cool in a

The leech, M. montezuma, feeds in the open water column of water bath, and diluted to 50.0 mL with distilled deionized water.

Montezuma Well at night where it senses its prey, H. montezuma, Trace-metal grade nitric acid was used for all digests and dilutions.

from the waves created by swimming. During the daytime, M. mon- A Perkin-Elmer Elan 500 ICP-MS with a glass nebulizer and

tezuma retreats to the sediments at the bottom of the well to rest water-cooled spray chamber was used to determine total arsenic

and wait for darkness. The insects that feed on H. montezuma live in in sediment, soil, and organisms. Sample was delivered to the neb-

−1

the water plants surrounding the edge of the Well and require day- ulizer by a peristaltic pump at a ﬂow rate of 1.5 mL min . Gallium

light to see their prey. To avoid predation, H. montezuma migrates was used as an internal standard and mass-to-charge (m/z) ratios

between the open water column during the day and the water were measured by peak hopping at 75 and 69. Calibration stan-

−1

plants at the edge of the Well at night [43]. This behavior is illus- dards, ranging from 0 to 1000 ␮g L , were prepared in 2% nitric

trated in Fig. 1, which shows the dynamics of the Montezuma Well acid by serial dilution of Spex plasma standards. A calibration blank

food web. and a calibration standard were run with every 10 samples. A lab

duplicate and a spike of the samples were also run with each 10

−1

samples. The calculated detection limit was 0.5 ␮g L , based on

2. Sample collection and total arsenic analysis

three times the standard deviation of the blanks. Table 4 sum-

marizes the reference materials used and the percent recoveries

Samples were collected from Montezuma Well in May and

obtained from these procedures.

September during the peak productivity season. The following sam-

ples were collected for analysis: H. montezuma, Hyalella azteca,

R. montezuma, B. bakeri, T. salva, M. montezuma, neuston (a 2.1. XANES data collection and analysis

phytoplankton assemblage) and sediment. B. bakeri, T. salva, R. mon-

tezuma and M. montezuma were collected by hand, frozen in liquid X-ray absorption near-edge spectroscopy (XANES) is a sensitive

nitrogen at the site and transported to the laboratory in individual tool for probing the local atomic and electronic environment of

centrifuge tubes. H. montezuma and H. azteca were collected in the an element, such as arsenic in a complex environment. It probes

water column with a 254-␮m pore-size plankton tow net. H. mon- all forms of arsenic in all phases (solid, aqueous, etc.) and can,

tezuma was sampled in the open water column, and H. azteca was with careful analysis, be used to quantify the different forms of

collected along the shoreline. M. montezuma was collected at dusk arsenic present in a particular sample [44,45]. Arsenic K-edge

in the open water column using a hand-operated dip net. Neuston X-ray absorption spectroscopy was carried out on beamline 7–3

was collected by skimming the water surface with a plankton hand of the Stanford Synchrotron Radiation Laboratory using a Si(2 2 0)

Table 4

Standard reference materials used for arsenic analysis.

−1

Standard reference material Sample matrix represented Certiﬁed arsenic value (mg kg ) Percent recovery

Estuarine sediment, NIST 1646 Sediment 11.6 ± 1.3 94 ± 6 (n = 6)

Montana soil, NIST 2711 Soil 105 ± 8 113 ± 4 (n = 4)

Dogﬁsh tissue, DORM-2 NCR Invertebrate 18.0 ± 1.1 99 ± 2 (n = 6)

562 R.D. Foust Jr. et al. / Coordination Chemistry Reviews 306 (2016) 558–565

3.5 As(V)

3.0 As(III)

(GS ) As(III) 2.5 3 AsB 2.0

1.5

1.0

0.5

Normalized Absorbance 0.0

118 60 11870 11880 11890

Energy (eV)

Fig. 2. As K-edge X-ray absorption spectra of model arsenic species.

double crystal monochromator, an upstream vertical aperture of

1 mm, and no focusing optics. Incident and transmitted intensities

were measured using N2-ﬁlled ion chambers, and the X-ray

absorption spectrum of the sample was recorded by monitoring

␣

As K ﬂuorescence using a Canberra 30-element germanium

detector. During data collection, samples were maintained at

approximately 10 K in a ﬂowing liquid helium cryostat. A spectrum

of elemental arsenic was collected simultaneously with that of

each sample in order to calibrate the energy scale; the ﬁrst energy

inﬂection of the As(0) was taken as 11,867.0 eV. Data collection

was carried out using the program XAS COLLECT [46], and data

reduction was carried out using the EXAFSPAK suite of programs

(http://ssrl.slac.stanford.edu/exafspak.html) according to standard

methods.

Near-edge spectra were ﬁt to a linear combination of spectra

from arsenic model compounds in a least squares procedure using

the program DATFIT in EXAFSPAK. All spectra are normalized, and

hence, the fractional contribution of the arsenic model spectrum to

the ﬁt is quantitatively equivalent to the fraction of arsenic present

as that type of species in the sample under scrutiny. Analytical stan-

dards were purchased from Sigma–Aldrich and were of the highest

quality available. Model compounds for ﬁtting were selected to

represent the types of arsenic species expected in a freshwater

environment and included the following: aqueous solutions of

arsenate (pH 9) and arsenite (pH 5.5), representing inorganic As(V)

and As(III); and As(III)-tris-glutathione [(GS)3As(III)] and arsenobe-

taine, representing organic compounds with arsenic coordinated

to sulfur and methyl groups, respectively. Solutions of arsenic

model compounds had a concentration of between 5.0 and 7.5 mM

after the addition of 30% glycerol as a glassing agent. Arsenic (III)-

tris-glutathione complex was made by adding a 10-fold excess

of glutathione to a solution of sodium arsenite to give a pH 5.5

solution [44,47]. Dimethyl arsenate, DMA(V), commonly known as

cacodylate) was also tested, but gave inferior ﬁts to those using

only arsenobetaine. Aqueous solutions were pipetted into 2-mm

path-length Lucite sample holders and frozen in liquid nitrogen.

Fig. 1. Dynamics of the Montezuma Well food web. The endemic fresh water

The model arsenic spectra should be considered to be representa-

amphipod Hyalella montezuma feeds on phytoplankton ﬂoating freely in the main

tive of a type of environment, rather than a speciﬁc molecule. For

water column of Montezuma Well. H. montezuma is the exclusive food source for

example, the determination of a fraction of (GS) As(III) indicates

the endemic leech, Motobdella montezuma and four insects, Cybister ellipticus (not 3

shown), Belostoma bakeri, Ranatra montezuma and Telebasis salva. The leech feeds As(III) coordinated by sulfur, which may or may not be from glu-

in the open water column at night, and the insects feed during the day in the weed

tathione. Component contributions were set to zero in the ﬁnal ﬁt

beds along the shore. This feeding behavior leads to a daily migration for the amphi-

if they contributed <1% of the ﬁt.

pod, which spends nights in the weeds along the shore and days in the open water

column in the center of the well [43].

3. Results and discussion

Fig. 2 shows the near-edge spectra of select arsenic models,

demonstrating the sensitivity of the As K-edge not only to oxidation

R.D. Foust Jr. et al. / Coordination Chemistry Reviews 306 (2016) 558–565 563

Total arsenic content in Montezuma Well organisms gener-

Neuston

ally decreases between successive trophic levels, a pattern that

is consistent with the results from previous studies [35,38]. The

3 Amphipo d

phytoplankton assemblage had a total arsenic concentration of

−1

702 mg kg , and the total arsenic content for the second trophic

Leech −1

level, the freshwater amphipod H. montezuma, is 3.4 mg kg , a 200-

2 fold decrease. The littoral zone insects, R. montezuma, B. bakeri and

−1

T. salva have arsenic levels ranging from 1.3 to <0.5 mg kg , rep-

resenting approximately an order of magnitude decrease between

1 the second and third trophic levels.

The extremely high total arsenic content of M. montezuma, the

endemic leech provides a notable exception to the overall diminu-

−1

tion trend. The measured 2810 mg kg is, to our knowledge, the

Normalized Absorbance 0

highest arsenic level reported in any organism. Earthworms col-

11850 11860 11870 11880 11890 11900 11910 lected from contaminated mining sites [55–57] are reported to have

Energy (eV) the highest arsenic levels for living organisms, with one specimen,

Lumbricus castaneous, from an abandoned gold mine in Nova Scotia,

−1

Canada reported to have an arsenic body burden of 2200 mg kg

Fig. 3. As K-edge X-ray absorption spectra of food web organisms.

[56].

The leech feed exclusively on the amphipod, H. montezuma. Most

state but also to chemical environment. The pronounced peak on leech specimens we collected contained noticeable bulges, which

top of each of the edges is due to dipole-allowed 1s → 4p transitions. upon further examination yielded carcasses of recently consumed

XANES spectra for three food web components, neuston (a phy- amphipods. The only other arsenic sources for the leech would

−1

toplankton assemblage), H. montezuma and M. montezuma have be the water (110 ␮g L , 100% H3AsO4) or the sediments where

−1

strikingly diverse spectra (Fig. 3). Sample spectra were ﬁt by a linear the leech resides at night (427 mg L ). XANES spectra show 100%

combination of model spectra to determine the % abundance of each of the leech arsenic is sulfur-coordinated. There is no evidence

type of arsenic species in the sample. The precision of the model for arsenate or non-sulfur containing organoarsenic complexes in

compound contributions, expressed as three times the estimated the XANES spectrum. M. montezuma uses enzymes to digest the

standard deviation, was <5% for all models and samples. Overall internal ﬂuids of H. montezuma and excrete the empty carcasses

model ﬁt residuals, expressed as 100 × (Io − Ic) /N, where Io and [43] and then appears to store dietary arsenic as a glutathione-

Ic are the observed and calculated intensities, respectively, and like (i.e., sulfur-coordinated) complex. In a separate experiment,

the summation is over the number of data points, N, were <0.136, two leech specimens were washed with a 0.1 M sodium phos-

except for the insect samples, which ranged from 0.298 to 1.70. phate solution and re-analyzed for total arsenic. Total arsenic in

−1

The phytoplankton assemblage (neuston), at the bottom of the the washed samples dropped to 45.5 mg kg , suggesting that most

food web, shows the most oxidized spectrum (80%). Although the of the arsenic–sulfur complex was on the surface of the organism

phytoplankton spectrum is ﬁt well by the aqueous arsenate model and not located within the tissue. This observation suggests that

in the near-edge region, the spectra do not correspond well in the leech may be transforming arsenic from its diet to a sulfur-

the post-edge region (11880–11900 eV) [45,48,49]. Variations in coordinated complex and excreting it to the membrane on its

this part of the spectrum are indicative of changes in longer-range surface (which is visible by microscope). This may indicate that

order [50], and the arsenate in these samples may be adsorbed to the biotransformed As is stored in this manner as a detoxiﬁcation

a mineral surface or coordinated in some other way. mechanism, a possibility that will be examined in future work.

The amphipod spectra differed from the phytoplankton in show- The only arsenic source for Montezuma Well is the feed water

ing a signiﬁcant (∼1/3) contribution of methylated arsenic, with that enters from vents at the bottom of the water column, and it was

the remainder being predominantly inorganic As(V) and sulfur- shown previously that the only form of arsenic in the water is arse-

coordinated As(III). As might be expected, the two amphipod nate, H3AsO4 [40]. Many studies have shown that phytoplankton

species analyzed (H. montezuma and H. azteca) showed arsenic dis- ingests arsenic from water, consistent with our data [33–36,38,58].

tributions that were substantially similar (data not shown). The Other studies report the form of arsenic in phytoplankton is pre-

insect samples had very low total arsenic levels and consequently, dominantly a mixture of arsenate with a small percentage as

XANES spectra for these samples suffered from noisier spectra and arsenite. MMA(V), DMA(V) and arsenobetaine have been reported

higher ﬁt residuals. Although the lower quality data limit the reli- in phytoplankton, but always at much smaller percentages than

ability of the exact numerical abundances of a given arsenic model inorganic arsenic [59]. The Montezuma Well phytoplankton assem-

within these samples, it still allows a general assessment of the blage shows a similar pattern.

relative presence reduced and oxidized forms. The lower trophic The endemic amphipod, H. montezuma contains less arsenic

−1

R. montezuma contains a greater proportion of oxidized arsenic than the phytoplankton upon which it feeds, 702 mg kg As for

∼ −1

( 65%) than the higher trophic B. bakeri (∼26%) and T. salva (∼41%). the phytoplankton assemblage compared to 3.4 mg kg As for

Finally, the top predator, the M. montezuma leech, contains arsenic the amphipod. The percentage of inorganic arsenic is signiﬁcantly

solely in the reduced and sulfur-coordinated form. reduced between trophic levels as well from 86 to 43%. Three

Total arsenic concentration data for food web organisms [51] insects living in the aquatic plants around the edge of the well,

are shown with the speciation data in Fig. 4, which illustrates the the littoral zone, constitute the third trophic level, in addition

transfer and transformation of arsenic through the Montezuma to the leech mentioned previously, and they feed exclusively on

Well food web. Based on the trophic structure and productivity this endemic amphipod. Total arsenic levels show approximately

model developed by Blinn and co-workers [42,52–54], we have an order of magnitude decrease from the second trophic level,

identiﬁed the quantity and the speciation type of arsenic for >90% and the arsenic speciation varied considerably within these three

of the biomass in the Montezuma Well food web. This is possible organisms. The fact that dietary arsenic from the same food source

because Montezuma Well is a simple closed system and the feeding leads to different speciation between these three insect species

characteristics of the endemic species are well known. suggests that ingested arsenic is metabolized differently, i.e.

564 R.D. Foust Jr. et al. / Coordination Chemistry Reviews 306 (2016) 558–565

Littoral Zone Limnetic Zone Belostma bakeri Telebasis salva Motobdella montezuma

-1 -1

(0.6 mg kg -1) (<0.5 mg kg ) (2,810 mg kg )

As(V ) 26%, As(III) 0%, As(V) 41%, As(III) 0%, As(V) 0%, As(III) 0%,

(GS) As(III) 0%, AsB 59% (GS) As(III) 100%, AsB 0% (GS)3As(III) 17%, AsB 57% 3 3

Ran atra montezuma

(1.3 mg kg -1)

As(V ) 65%, As(III) 0%,

(GS)3As(III) 23%, AsB 12% Hyalella montezuma (3.4 mg kg -1)

As(V) 35%, As(III) 8%,

(GS)3As(III) 22%, AsB 35%

Prima ry Producers

(702 mg kg -1)

As(V ) 80%, As(III) 6%,

(GS)3As(III) 11%, AsB 3%

Montezuma Well sediment Montezuma Well ambient water

(427 mg kg -1) (109 µg L -1)

As(V) 14 %, As(III) 0%, As(V) 100%, As(III) 0%,

(GS)3As(III) 74%, DMA 11% (GS)3As(III) 0%, AsB 0%

Fig. 4. Total arsenic concentrations determined by ICP-MS and relative abundance of arsenic species types (expressed as model compounds) determined by XANES for the

organisms in the Montezuma Well food web. This ﬁgure is adapted from a model for the trophic structure, production, and energy ﬂow in the littoral and limnetic zones in

Montezuma Well [42,53,54]. Because of the high XANES ﬁt residuals for the insects, speciation abundances are for informational purposes only. Total arsenic concentrations

are reported as dry weight.

arsenic speciation is species-speciﬁc [60]. It is likely that the As(V) for Research Resources, Biomedical Technology Program, and the

observed for the three insects was arsenate adsorbed to the surface National Institute of General Medical Sciences. We are grateful to

of their exoskeleton which could have been picked up from the Mr. Glenn Henderson and Ms. Kathy Davis, Park Superintendents

water in which they live. for Montezuma Well National Monument, for their support and

for granting access to the Well as needed for this research project.

4. Conclusions We thank Professor Michael Ketterer, Metropolitan State Univer-

sity of Denver, for assistance obtaining ICP-MS data. Finally, we

We have shown for the ﬁrst time how arsenic is distributed in wish to thank Professor David Salt, Purdue University, for assistance

>90% of the biomass of a freshwater food web. Our data show a in gaining access to the facilities at SSRL and for his many helpful

general biodiminution of arsenic in a freshwater food web, even suggestions in preparing biological samples for XANES analysis.

as the top predator, M. montezuma, has a higher arsenic con-

tent than any previously reported organism. The biotransformed References

(sulfur-coordinated) arsenic in M. montezuma appears to be present

[1] J.M. Azcue, J.O. Nriagu, in: J.O. Nriagu (Ed.), Arsenic in the Environment Part I:

almost entirely on the surface of the organism, however, suggesting

Cycling and Characterization, John Wiley & Sons, Inc., New York, USA, 1994, pp.

a detoxiﬁcation process. Our insect data illustrate that the transfor- 1–16.

mations of arsenic in the Montezuma Well ecosystem appear to be [2] B.K. Mandal, K.T. Suzuki, Talanta 58 (2002) 201–235.

species-speciﬁc. [3] U.S. Geological Survey, Mineral Commodity Summaries 2013, U.S. Geological

Survey, Reston, Virginia, 2013.

[4] M.S. Gorby, in: J.O. Nriagu (Ed.), Arsenic in the Environment, Part II: Human

Acknowledgments Health and Ecosystem Affects, John Wiley & Sons, Inc., New York, 1994,

pp. 1–16.

[5] M.F. Hughes, B.D. Beck, Y. Chen, A.S. Lewis, D.J. Thomas, Toxicol. Sci. 123 (2011)

We acknowledge the ﬁnancial assistance of the U.S. Depart- 305–352.

ment of Energy through two cooperative agreements, DE- [6] K.J. Williams, J. R. Soc. Med. 102 (2009) 343–349.

[7] J.O. Nriagu, in: W.T. Frankenberger Jr. (Ed.), Environmental Chemistry of

FC04-90AL66158 and DE-FC02-02-EW15254, administered by

Arsenic, Marcel Dekker, New York, 2002.

the HBCU-MI Environmental Technology Consortium. We also

[8] F.J. Peryea, 16th World Congress of Soil Science, Montpellier, France, 1998.

acknowledge the ﬁnancial assistance of the National Science Foun- [9] T. Schooley, M.J. Weaver, D. Mullins, M. Eick, J. Pestic. Saf. Educ. 10 (2008)

dation for grants DBI-0244221 and CHE-0116804. This work was 22–32.

[10] J.O. Nriagu, J.M. Azcue, in: J.O Nriagu, M.S. Simmons (Eds.), Food Contamination

performed at SLAC National Accelerator Laboratory, Menlo Park, CA,

from Environmental Sources, John Wiley & Sons, Inc., New York, USA, 1990,

operated by Stanford University for the U.S. Department of Energy

pp. 121–144.

Ofﬁce of Science. The Stanford Synchrotron Radiation Laboratory [11] A. Chatterjee, D. Das, B.K. Mandal, T.R. Chowdhury, G. Samanta, D. Chakraborti,

Analyst 120 (1995) 643–650.

is funded by the Department of Energy, Ofﬁces of Basic Energy

[12] T.R. Chowdhury, G.K. Basu, B.K. Mandal, B.K. Biswas, G. Samanta, U.K. Chowd-

Sciences and Biological and Environmental Research (DE-AC03-

hury, C.R. Chanda, D. Lodh, S.L. Roy, K.C. Saha, S. Kabir, Q. Quamruzzaman, D.

76SF00515); the National Institutes of Health, National Center Chakraborti, Nature 401 (1999) 545–547.

R.D. Foust Jr. et al. / Coordination Chemistry Reviews 306 (2016) 558–565 565

[13] A.B. Mukherjee, P. Bhattacharya, Environ. Rev. 9 (2001) 189–220. [38] G. Caumette, I. Koch, E. Estrada, K.J. Reimer, Environ. Sci. Technol. 45 (2011)

[14] E. Lombi, K.G. Scheckel, J. Pallon, A.M. Carey, Y.G. Zhu, A.A. Meharg, New Phytol. 9917–9923.

184 (2009) 193–201. [39] A.D. Konieczki, S.A. Keake, Hydrogeology and Water Chemistry of Montezuma

[15] P.N. Williams, A.H. Price, A. Raab, S.A. Hossain, J. Feldmann, A.A. Meharg, Envi- Well and Montezuma Castle National Monument and Surrounding Area, Ari-

ron. Sci. Technol. 39 (2005) 5531–5540. zona, U.S. Department of the Interior, Tucson, AZ, 1997.

[16] B. Burton-Freeman, L. Jackson, J. Johnston, J. Maclachlan, W.R. Mindak, 245th [40] R.D. Foust Jr., P. Mohapatra, A.M. Compton-O’Brien, J. Reifel, Appl. Geochem. 19

ACS National Meeting, American Chemical Society, New Orleans, LA, 2013. (2004) 251–255.

[17] D.T. Heitkemper, K.M. Kubachka, P.R. Halpin, M.N. Allen, N.V. Shockey, Food [41] R.D. Foust Jr., M. Brandstrom, G. Evans, P. Nez, D. Waupoose, E. Hillman, Arsenic

Addit. Contam. B: Surveill. 2 (2009) 112–120. in Soil and Groundwater Environment, Elsevier Ltd., 2007, pp. 85–100.

[18] J.G. Hering, P.E. Kneebone, in: W.T. Frankenberger Jr. (Ed.), Environmental [42] C. Runck, D.W. Blinn, Limnol. Oceanogr. 39 (1994) 1800–1812.

Chemistry of Arsenic, Marcel Dekker, New York, 2002, pp. 155–161. [43] D.W. Blinn, in: L.E Stevens, V.J. Meretsky (Eds.), Aridlands Springs in North

[19] J. Matschullat, Sci. Total Environ. 249 (2000) 297–312. America: Ecology and Conservation, University of Arizona Press, Tucson, 2008,

[20] P.L. Smedley, D.G. Kinniburgh, Appl. Geochem. 17 (2002) 517–568. pp. 98–126.

[21] K.J. Reimer, I. Koch, W.R. Cullen, in: A. Sigel, H. Sigel, R.K.O. Sigel (Eds.), [44] G.N. George, R.C. Prince, S.P. Singh, I.J. Pickering, Mol. Nutr. Food Res. 53 (2009)

Organometallics in Environment and Toxicology: Metal Ions in Life Sciences, 552–557.

Royal Society of Chemistry, 2010, pp. 165–229. [45] I.J. Pickering, G.N. George, AIP Conf. Proc. 882 (2007) 311–315.

[22] E. Dopp, A.D. Kligerman, R.A. Diaz-Bone, in: A. Sigel, H. Sigel, R.K.O. Sigel (Eds.), [46] M.J. George, J. Synchrotron Radiat. 7 (2000) 283–286.

Organometallics in Environment and Toxicology: Metal Ions in Life Sciences, [47] I.J. Pickering, R.C. Prince, M.J. George, R.D. Smith, G.N. George, D.E. Salt, Plant

Royal Society of Chemistry, 2010, pp. 231–265. Physiol. 122 (2000) 1171–1177.

[23] K.A. Francesconi, D. Kuehnelt, in: W.T. Frankenberger Jr. (Ed.), Environmental [48] J.J.H. Cotelesage, M.J. Pushie, P. Grochulski, I.J. Pickering, G.N. George, J. Inorg.

Chemistry of Arsenic, Marcel Dekker, New York, 2002, pp. 51–94. Biochem. 115 (2012) 127–137.

[24] B.M. Adair, S.B. Waters, V. Devesa, Z. Drobna, M. Styblo, D.J. Thomas, Environ. [49] G.N. George, I.J. Pickering, M.J. Pushie, K. Nienaber, M.J. Hackett, I. Ascone, B.

Chem. 2 (2005) 161–166. Hedman, K.O. Hodgson, J.B. Aitken, A. Levina, C. Glover, P.A. Lay, J. Synchrotron

[25] R.S. Oremland, J.F. Stolz, Science 300 (2003) 939–944. Radiat. 19 (2012) 875–886.

[26] M.A. Rahman, H. Hasegawa, R.P. Lim, Environ. Res. 116 (2012) 118–135. [50] S.M. Webb, J.F. Gaillard, L.Q. Ma, C. Tu, Environ. Sci. Technol. 37 (2003) 754–760.

[27] R. Bentley, T.G. Chasteen, Microbiol. Mol. Biol. Rev. 66 (2002) 250–271. [51] A.-M. Compton-O’Brien, R.D. Foust Jr., M.E. Ketterer, D.W. Blinn, in: Y Cai, O.C.

[28] F. Challenger, Chem. Rev. 36 (1945) 315–361. Braids (Eds.), Biogeochemistry of Environmentally Important Trace Elements,

[29] F. Challenger, Adv. Enzymol. Relat. Areas Mol. Biol. 12 (1951) 429–491. American Chemical Society, Washington, DC, 2003, pp. 200–209.

[30] M. Vahter, Toxicology 181–182 (2002) 211–217. [52] B. Dehdashti, D.W. Blinn, Freshw. Biol. 25 (1991) 131–141.

[31] H. Hasegawa, Y. Sohrin, K. Seki, M. Sato, K. Norisuye, K. Naito, K. Naito, M. Matsui, [53] C. Runck, D.W. Blinn, J. N. Am. Benthol. Soc. 9 (1990) 262–270.

Chemosphere 43 (2001) 265–272. [54] C. Runck, D.W. Blinn, J. N. Am. Benthol. Soc. 12 (1993) 136–147.

[32] W.T. Frankenberger Jr., M. Arshad, in: W.T. Frankenberger Jr. (Ed.), Environ- [55] M. Button, G.R.T. Jenkin, C.F. Harrington, M.J. Watts, J. Environ. Monit. 11 (2009)

mental Chemistry of Arsenic, Marcel Dekker, New York, 2002, pp. 363–381. 1484–1491.

[33] J.M. Giddings, G.K. Eddlemon, Arch. Environ. Contam. Toxicol. 6 (1977) [56] M. Button, I. Koch, K.J. Reimer, Environ. Pollut. 169 (2012) 74–80.

491–505. [57] M. Button, M.M. Moriarty, M.J. Watts, J. Zhang, I. Koch, K.J. Reimer, Chemosphere

[34] D.M. Lindsay, J.G. Sanders, Environ. Toxicol. Chem. 9 (1990) 391–395. 85 (2011) 1277–1283.

[35] C.Y. Chen, C.L. Folt, Environ. Sci. Technol. 34 (2000) 3878–3884. [58] S. Maeda, K. Kusadome, H. Arima, A. Ohki, K. Naka, Appl. Organomet. Chem. 6

[36] S. Maeda, R. Inoue, T. Kozono, T. Tokuda, A. Ohki, T. Takeshita, Chemosphere 20 (1992) 399–405.

(1990) 101–108. [59] S. Maeda, in: J.O. Nriagu (Ed.), Arsenic in the Environment, Part I: Cycling and

[37] T. Kuroiwa, A. Ohki, K. Naka, S. Maeda, Appl. Organomet. Chem. 8 (1994) Characterization, John Wiley & Sons, Inc., New York, USA, 1994, pp. 155–188.

325–333. [60] M.A. Rahman, C. Hassler, Aquat. Toxicol. 146 (2014) 212–219.